Inhibiting the effectiveness of students' inquiry learning, our research (Kuhn et al., 2001) suggests, is an immature "co-occurrence" mental model of causality. A simple co-occurrence of one level of the variable with an outcome is sufficient to implicate that variable in the outcome. The potential causal influence of a second variable, therefore, need not be treated as additive. Instead, the second variable can be invoked as a different explanation for a later outcome, or the second variable can be discounted because the first feature explains the outcome. Accordingly, the co-occurrence mental model treats causal influences as neither consistent nor additive. When multiple variables co-occur with variation in an outcome, instead of recognizing the indeterminacy with respect to cause, students with weak analysis strategies may focus the explanatory burden on only those variable(s) they already believe to be causal and shift from one to another in an inconsistent manner. In the earthquake problem, for example, a student may attribute one case of high risk to a high level of snake activity, and then go on to attribute a case having low risk to type of bedrock, this time ignoring the high snake activity that continues to be present, and in both cases ignoring other variables that have also co-varied with outcome.
Kuhn, D., Black, J., Keselman, A., & Kaplan, D. (2000). The development of cognitive skills to support inquiry learning. Cognition and Instruction, 18, 495-523.